On Thu, 24 Sep 2026 16:09:39 +0200 Paolo Abeni wrote:
> I'm wondering if this patch is needed at all. Can't the special gso caller
> set skb_shinfo(skb)->gso_size to ~64K and adjust the gso bits in the shared
> info afterwards?

That would work, and it keeps the logic in one place without a new cb field,
but I don't think it's better than the current approach.

It would need skb_unclone() before writing skb_shinfo(), and a repeat of the
if (partial_segs) logic from skb_segment().  For DF=0 we would also have to
recompute ip.id in validate_xmit_skb(), or keep that case on the existing
full-segmentation path.
_______________________________________________
dev mailing list
[email protected]
https://mail.openvswitch.org/mailman/listinfo/ovs-dev

Reply via email to